CN114720436B - Agricultural product quality parameter detection method and equipment based on fluorescence hyperspectral imaging - Google Patents
Agricultural product quality parameter detection method and equipment based on fluorescence hyperspectral imaging Download PDFInfo
- Publication number
- CN114720436B CN114720436B CN202210077713.2A CN202210077713A CN114720436B CN 114720436 B CN114720436 B CN 114720436B CN 202210077713 A CN202210077713 A CN 202210077713A CN 114720436 B CN114720436 B CN 114720436B
- Authority
- CN
- China
- Prior art keywords
- vissa
- ivmr
- iriv
- agricultural products
- fluorescence hyperspectral
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/62—Systems in which the material investigated is excited whereby it emits light or causes a change in wavelength of the incident light
- G01N21/63—Systems in which the material investigated is excited whereby it emits light or causes a change in wavelength of the incident light optically excited
- G01N21/64—Fluorescence; Phosphorescence
- G01N21/6402—Atomic fluorescence; Laser induced fluorescence
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A40/00—Adaptation technologies in agriculture, forestry, livestock or agroalimentary production
- Y02A40/10—Adaptation technologies in agriculture, forestry, livestock or agroalimentary production in agriculture
Landscapes
- Health & Medical Sciences (AREA)
- Physics & Mathematics (AREA)
- Analytical Chemistry (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Optics & Photonics (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Investigating, Analyzing Materials By Fluorescence Or Luminescence (AREA)
Abstract
The invention discloses a method and equipment for detecting quality parameters of agricultural products based on fluorescence hyperspectral imaging, belonging to the field of agricultural product detection, and comprising the following steps: s1, collecting fluorescence hyperspectral image data; s2, preprocessing the acquired data; s3, performing feature extraction on the spectral image by using a IRIV, VISSA, MASS and RF algorithm, directly combining the extracted features to obtain an IVMR combined feature variable, performing feature extraction by using a VISSA algorithm to obtain an IVMR-VISSA feature variable, and performing feature extraction by using an IRIV algorithm to obtain an IVMR-VISSA-IRIV feature variable; s4, based on IVMR-VISSA-IRIV characteristic variables as input, constructing an MK-SVR model; s5, predicting the internal quality parameters of the agricultural products by using an MK-SVR model. The method has high accuracy and provides a new technical idea for nondestructive testing analysis of quality parameters of agricultural products.
Description
Technical Field
The invention relates to the technical field of agricultural product detection, in particular to an agricultural product quality parameter detection method and equipment based on fluorescence hyperspectral imaging.
Background
Nondestructive testing of internal quality parameters such as Soluble Solids Content (SSC), PH, moisture, and vitamin content of agricultural products (e.g., kiwi fruit, etc.) is increasingly being studied. At present, a common detection means of internal quality parameters is a lossy physicochemical experiment, and the result is accurate, but time and labor are wasted. In the prior art, nondestructive detection of fruits is performed by utilizing a visible/near infrared hyperspectral imaging technology, but the detection by utilizing a fluorescence hyperspectral imaging technology is rarely performed. On the other hand, there is a scheme of processing spectral data by an algorithm, but there is a problem of poor prediction accuracy.
Disclosure of Invention
The invention aims to overcome the defects of the prior art, provides a method and equipment for detecting quality parameters of agricultural products based on fluorescence hyperspectral imaging, has high accuracy, and provides a new technical idea for nondestructive detection analysis of the quality parameters.
The invention aims at realizing the following scheme:
a method for detecting quality parameters of agricultural products based on fluorescence hyperspectral imaging comprises the following steps:
s1, collecting fluorescence hyperspectral image data of agricultural products;
s2, preprocessing the collected fluorescence hyperspectral image data of the agricultural products;
s3, performing feature extraction on the spectral image preprocessed in the step S2 by adopting IRIV, VISSA, MASS and RF algorithms respectively to obtain four corresponding feature variables, then directly combining the four feature variables to obtain an IVMR combined feature variable, performing feature extraction on the IVMR combined feature variable by using a VISSA algorithm to obtain an IVMR-VISSA feature variable, and performing feature extraction on the IVMR-VISSA feature by using an IRIV algorithm to obtain an IVMR-VISSA-IRIV feature variable;
s4, taking the IVMR-VISSA-IRIV characteristic variable obtained in the step S3 as input, taking the physicochemical value of the agricultural product quality parameter as output to form a training sample pair, training a multi-core support vector regression model MK-SVR, setting polynomial kernel functions with different orders in the training process, and selecting a decision coefficient R C 2 The penalty coefficient c corresponding to the maximum value is taken as the optimal penalty coefficient c best Thereby establishing an MK-SVR prediction model;
s5, obtaining the corresponding IVMR-VISSA-IRIV characteristic variable of the agricultural product to be tested through the steps S1 to S3, inputting the IVMR-VISSA-IRIV characteristic variable into the trained MK-SVR model obtained in the step S4 for prediction, and obtaining the internal quality parameter of the agricultural product to be tested.
Further, in step S1, the fluorescence hyperspectral image data includes average spectral data of a region of interest selected from the fluorescence hyperspectral image.
Further, in step S2, the preprocessing includes preprocessing the fluorescence hyperspectral image data of the agricultural product acquired in step S1 by using a Detrending algorithm.
Further, in step S3, in step S4, the establishing the multi-core support vector regression prediction model MK-SVR includes the sub-steps of:
selecting a polynomial function as a basic kernel function, and performing weighted linear combination on a plurality of basic kernel functions to obtain a multi-core kernel function, wherein the multi-core kernel function is calculated as follows:
in the above, k m (x, y) is the mth mononuclear function, θ m Is the weight of the mth mononuclear function.
Further, the plurality of basic kernel functions includes polynomial kernel functions of five different orders d.
Further, the agricultural product includes kiwi fruit.
Further, the average spectrum data of the region of interest is specifically obtained by using ENVI software to take a selected region in the fluorescence hyperspectral image of the agricultural product as the region of interest, and calculating the average spectrum data of the region.
In step S2, the Detrending algorithm Detrending is used to remove noise on the fluorescence hyperspectral image.
Further, the selected area is 3/4.
A computer device comprising a processor and a memory, the memory having stored therein a computer program which, when loaded by the processor, performs the method of any of the preceding claims.
The beneficial effects of the invention include:
the method provided by the embodiment of the invention verifies the feasibility of predicting the PH content of the kiwi fruits by adopting a fluorescence hyperspectral imaging technology, designs primary, secondary and tertiary feature extraction methods, establishes corresponding prediction models, and provides technical analysis and method reference for nondestructive detection of the PH content of the kiwi fruits by using the fluorescence hyperspectral imaging technology.
In the embodiment of the invention, the Detrending algorithm is selected to preprocess the fluorescence hyperspectral image data of the agricultural products, so that noise in the fluorescence hyperspectral image data can be removed, and IVMR combination characteristic variables are obtained by combining after primary characteristic extraction; further, performing secondary feature extraction and tertiary feature extraction on the IVMR combined feature variables respectively and sequentially through VISSA and IRIV to further remove redundant variables, and using MK-SVR as a regression prediction model. Proved by verification, the IVMR-VISSA-IRIV-MK-SVR has optimal prediction effect, R P 2 、R C 2 RPD 0.8512, 0.8580 and 2.66, respectively. Therefore, the method provided by the embodiment of the invention is not only feasible, but also has higher accuracy, and provides a new technical idea for the nondestructive testing analysis of the PH content of the kiwi fruits and even other agricultural product quality parameters in the future.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions of the prior art, the drawings which are used in the description of the embodiments or the prior art will be briefly described, it being obvious that the drawings in the description below are only some embodiments of the invention, and that other drawings can be obtained according to these drawings without inventive faculty for a person skilled in the art.
FIG. 1 is a fluorescence hyperspectral image of kiwi fruit;
FIG. 2 is a spectral image after Detrending pretreatment;
FIG. 3 is a characteristic variable distribution of IRIV extraction;
FIG. 4 is a distribution of feature variables extracted by VISSA;
FIG. 5 is a distribution of feature variables extracted by MASS;
FIG. 6 is a characteristic variable distribution of RF extraction;
FIG. 7 is a characteristic variable distribution of IVMR-VISSA extraction;
IVMR-VISSA, IVMR-VISSA and IVMR-VISSA-IRIV are self-naming terms.
Detailed Description
All of the features disclosed in all of the embodiments of this specification, or all of the steps in any method or process disclosed implicitly, except for the mutually exclusive features and/or steps, may be combined and/or expanded and substituted in any way.
The technical conception, the technical problems to be solved, the working principle, the working process and the beneficial effects of the invention are further and fully described below with reference to fig. 1 to 7.
1. Extraction of spectral data
See fig. 1. According to the embodiment of the invention, a xenon lamp can be used as an excitation light source, 390nm is selected as an excitation filter, 495nm is selected as a fluorescence filter, and a hyperspectral fluorescence system is utilized to obtain a fluorescence hyperspectral image of the kiwi fruit. And using ENVI software to take 3/4 area in the kiwi fruit fluorescence hyperspectral image as a region of interest (region of interest, ROI) and extract a spectrum, taking the average spectrum of the ROI area as fluorescence hyperspectral image data of a kiwi fruit sample, wherein the fluorescence hyperspectral image of the kiwi fruit sample is shown in figure 1. The wavelength range is 376.80 nm-1011.05 nm, and the total spectrum information of 125 wave bands is selected.
The 90 samples are divided into training sets and prediction sets by using an SPXY algorithm, wherein the training sets are 70, the prediction sets are 20, and the PH content statistical results of the divided kiwi fruits are shown in Table 1.
TABLE 1 Kiwi pH statistics
As can be seen from table 1, the PH range of the predicted set is within the PH range of the training set, and the dispersion degree of the samples of the training set and the predicted set is substantially consistent, which indicates that the samples of the two are representative.
2. Spectral pretreatment
See fig. 2. Noise generated in the process of extracting the spectral image can have a certain influence on a prediction result, so that fluorescence hyperspectral image data of the kiwi fruit is preprocessed by using a Detrending algorithm, and the preprocessed spectral image is shown in fig. 2.
3. Feature variable extraction
Performing primary feature extraction on the spectrum image subjected to Detrending pretreatment by adopting 4 algorithms such as IRIV, MASS, VISSA, RF and the like, directly combining 4 primary feature variables on the basis to obtain IVMR combined feature variables, and respectively performing VISSA secondary feature extraction and IRIV tertiary feature extraction on the IVMR combined feature variables, wherein the details are as follows:
referring to fig. 3, IRIV-based feature variable extraction: in the actual application process, extracting characteristic variables from the preprocessed spectrum data by adopting an iterative retention information variable method (Iteratively reserved information variables, IRIV). The maximum prime number of IRIV was set to 20 and the cross validation was 5 times. The distribution of IRIV extracted feature variables is shown in fig. 6, and 28 feature variables are extracted through 5 iterations.
Referring to fig. 4, VISSA-based feature variable extraction: in the practical application process, the preprocessed spectrum data is extracted into characteristic variables by adopting a variable iteration space contraction method (variable iterative space shrinkage approach, VISSA). The cross validation is set to be 5 folds, the binary matrix sampling number is 1000, the maximum main factor number is 10, and the initial weight of the variable is 0.5. Fig. 7 shows the distribution of feature variables extracted by VISSA, and 42 feature variables were extracted after 24 iterations.
Referring to fig. 5, MASS-based feature variable extraction: in the practical application process, the preprocessed spectrum data is extracted into characteristic variables through a model self-adaptive space reduction method (The Model Adaptive Space Shrinkage, MASS). The sample sampling rate is set to be 0.95, the characteristic sampling rate is set to be 0.5, the binary matrix sampling number is 1000, the maximum main factor number is 10, and the cross validation is 5 folds. The distribution of the feature variables extracted by the MASS is carried out for 25 iterations to extract 34 feature variables.
Referring to fig. 6, RF-based feature variable extraction: in the actual application process, the preprocessed spectrum data is extracted into characteristic variables by adopting a Random Frog-leaping algorithm (RF). The initial variable number is set to 5, iterated 2000 times, and the threshold probability is 0.15. FIG. 6 is a graph of the extracted feature variable distribution, extracting 33 feature variables with a probability of 0.15 or more.
In the embodiment of the invention, the characteristic variables extracted by one method are considered to contain limited information, so that 60 (IRIV+VISSA+MASS+RF and IVMR) combined characteristic variables obtained by combining the characteristic variables extracted by the 4 methods have more and correlative quantity, and secondary characteristic extraction is needed.
Referring to fig. 7, secondary feature extraction: in the actual application process, the VISSA algorithm is used for carrying out secondary feature extraction on the 60 IVMR combination feature variables. Setting the parameters as 5-fold cross validation, the binary matrix sampling number as 1000, the maximum main factor number as 10 and the variable initial weight as 0.5. The distribution of the characteristic variables extracted by the IVMR-VISSA is performed for 25 iterations to extract 27 characteristic variables. 27. Redundant variables are also present in the feature variables, so the embodiment of the invention performs three feature extraction on them.
Three times of feature extraction: three feature extraction were performed on 27 feature variables in IVMR-VISSA using IRIV. The maximum number of main factors was set to 20 and cross-validated 8 times. The distribution of the characteristic variables extracted by the IVMR-VISSA-IRIV is obtained by eliminating 4 characteristic variables such as 42, 1, 50, 43 and the like, and extracting 23 characteristic variables.
4. Modeling
And training a multi-core support vector regression model MK-SVR by using the IVMR-VISSA-IRIV characteristic variables extracted from the training set samples, and then inputting the IVMR-VISSA-IRIV characteristic variables of the prediction set into the trained MK-SVR model for prediction. The SVR model only sets one kernel function to predict the sample, and when the data source is wider, the prediction result is poorer. Considering that a plurality of kernel functions are mixed into one kernel function, the kernel function is called a multi-kernel function (Multiple Kernel Function), and the multi-kernel function can adapt to various characteristic variables to a greater extent, so that the accuracy of a prediction result can be improved.
The multi-core function is a weighted linear combination of multiple basic core functions that is calculated as follows:
wherein k is m (x, y) is the mth mononuclear function, θ m Is the weight of the mth mononuclear function.
The polynomial function is selected as a basic kernel function, and the polynomial kernel functions with five different orders are combined linearly to construct the multi-core kernel function. Setting polynomial order d= [1,2,3,4,5]The penalty coefficient C has a value range of [10 ] -1 ～10 6 ]The results of the MK-SVR model on the pH content of kiwi samples (including training set and predictive set) are shown in Table 2.
TABLE 2 MK-SVR prediction results for different feature extraction methods
In Table 2, R c 2 、R p 2 The decision coefficients of the training set and the prediction set are respectively represented (the closer the value is to 1, the higher the stability and the fitting degree of the model are represented), the RMSEC and the RMSEP are respectively represented by the root mean square error of the training set and the prediction set (the smaller the value is, the better the prediction performance of the model is represented by the smaller value), the RPD represents the relative analysis error (the higher the prediction accuracy of the model is represented by the RPD is more than or equal to 2), the difference of the prediction results of the IVMR-MK-SVR is derived from the fact that the redundancy among the IVMR combined feature variables is larger, the VISSA extracts the IVMR combined feature variables for the second time, the redundant variables are removed, the third feature extraction is further carried out by IRIV, and the prediction effect of the IVMR-VISSA-IRIV-MK-SVR is optimal, wherein R is P 2 、R C 2 RPD 0.8512, 0.8580 and 2.66, respectively.
Example 1: a method for detecting quality parameters of agricultural products based on fluorescence hyperspectral imaging comprises the following steps:
s1, collecting fluorescence hyperspectral image data of agricultural products;
s2, preprocessing the collected fluorescence hyperspectral image data of the agricultural products;
s3, performing feature extraction on the spectral image preprocessed in the step S2 by adopting IRIV, VISSA, MASS and RF algorithms respectively to obtain four corresponding feature variables, then directly combining the four feature variables to obtain an IVMR combined feature variable, performing feature extraction on the IVMR combined feature variable by using a VISSA algorithm to obtain an IVMR-VISSA feature variable, and performing feature extraction on the IVMR-VISSA feature by using an IRIV algorithm to obtain an IVMR-VISSA-IRIV feature variable;
s4, taking the IVMR-VISSA-IRIV characteristic variable obtained in the step S3 as input, taking the physicochemical value of the agricultural product quality parameter as output to form a training sample pair, training a multi-core support vector regression model MK-SVR, setting polynomial kernel functions with different orders in the training process, and selecting a decision coefficient R C 2 The penalty coefficient c corresponding to the maximum value is taken as the optimal penalty coefficient c best Thereby establishing an MK-SVR prediction model;
s5, obtaining the corresponding IVMR-VISSA-IRIV characteristic variable of the agricultural product to be tested through the steps S1 to S3, inputting the IVMR-VISSA-IRIV characteristic variable into the trained MK-SVR model obtained in the step S4 for prediction, and obtaining the internal quality parameter of the agricultural product to be tested.
Example 2: on the basis of the embodiment, in step S1, the fluorescence hyperspectral image data includes average spectral data of a region of interest selected from the fluorescence hyperspectral image.
Example 3: on the basis of embodiment 2, in step S2, the preprocessing includes preprocessing the fluorescence hyperspectral image data of the agricultural product acquired in step S1 by using Detrending algorithm.
Example 4: on the basis of embodiment 1, in step S3, in step S4, the building of the multi-core support vector regression prediction model includes the sub-steps of:
selecting a polynomial function as a basic kernel function, and performing weighted linear combination on a plurality of basic kernel functions to obtain a multi-core kernel function, wherein the multi-core kernel function is calculated as follows:
in the above, k m (x, y) is the mth mononuclear function, θ m Is the weight of the mth mononuclear function.
Example 5: on the basis of embodiment 4, the plurality of basic kernel functions includes polynomial kernel functions of five different orders d.
Example 6: on the basis of embodiment 1, the agricultural product comprises kiwi fruit.
Example 7: on the basis of the embodiment 2, the average spectrum data of the region of interest is specifically that the ENVI software is adopted to take the selected region in the fluorescence hyperspectral image of the agricultural product as the region of interest, and the average spectrum data of the region is calculated.
Example 8: on the basis of embodiment 3, in step S2, the Detrending algorithm Detrending is used to remove noise on the fluorescence hyperspectral image.
Example 9: based on embodiment 7, the selected area is 3/4.
Example 10: a computer device comprising a processor and a memory, the memory having stored therein a computer program which, when loaded by the processor, performs the method of any of embodiments 1-9.
The invention is not related in part to the same as or can be practiced with the prior art.
The foregoing technical solution is only one embodiment of the present invention, and various modifications and variations can be easily made by those skilled in the art based on the application methods and principles disclosed in the present invention, not limited to the methods described in the foregoing specific embodiments of the present invention, so that the foregoing description is only preferred and not in a limiting sense.
In addition to the foregoing examples, those skilled in the art will recognize from the foregoing disclosure that other embodiments can be made and in which various features of the embodiments can be interchanged or substituted, and that such modifications and changes can be made without departing from the spirit and scope of the invention as defined in the appended claims.
Claims (9)
1. The method for detecting the quality parameters of the agricultural products based on fluorescence hyperspectral imaging is characterized by comprising the following steps:
s1, collecting fluorescence hyperspectral image data of agricultural products;
s2, preprocessing the collected fluorescence hyperspectral image data of the agricultural products;
s3, performing feature extraction on the spectral image preprocessed in the step S2 by adopting IRIV, VISSA, MASS and RF algorithms respectively to obtain four corresponding feature variables, then directly combining the four feature variables to obtain an IVMR combined feature variable, performing feature extraction on the IVMR combined feature variable by using a VISSA algorithm to obtain an IVMR-VISSA feature variable, and performing feature extraction on the IVMR-VISSA feature by using an IRIV algorithm to obtain an IVMR-VISSA-IRIV feature variable;
s4, taking the IVMR-VISSA-IRIV characteristic variable obtained in the step S3 as input, taking the physicochemical value of the agricultural product quality parameter as output to form a training sample pair, training a multi-core support vector regression model MK-SVR, setting polynomial kernel functions with different orders in the training process, and selecting a decision coefficient R C 2 The penalty coefficient c corresponding to the maximum value is taken as the optimal penalty coefficient c best Thereby establishing an MK-SVR prediction model; in step S4, the building of the MK-SVR prediction model includes the sub-steps of: selecting a polynomial function as a basic kernel function, and performing weighted linear combination on a plurality of basic kernel functions to obtain a multi-core kernel function, wherein the multi-core kernel function is calculated as follows:
in the above, k m (x, y) is the mth mononuclear function, θ m Weights for the mth mononuclear kernel;
s5, obtaining the corresponding IVMR-VISSA-IRIV characteristic variable of the agricultural product to be tested through the steps S1 to S3, inputting the IVMR-VISSA-IRIV characteristic variable into the trained MK-SVR model obtained in the step S4 for prediction, and obtaining the internal quality parameter of the agricultural product to be tested.
2. The method for detecting quality parameters of agricultural products based on fluorescence hyperspectral imaging according to claim 1 wherein in step S1, the fluorescence hyperspectral image data includes average spectral data of a region of interest selected from fluorescence hyperspectral images.
3. The method for detecting quality parameters of agricultural products based on fluorescence hyperspectral imaging according to claim 1, wherein in step S2, the preprocessing includes preprocessing the fluorescence hyperspectral image data of the agricultural products collected in step S1 by using a Detrending algorithm.
4. The method for detecting quality parameters of agricultural products based on fluorescence hyperspectral imaging according to claim 1 wherein the plurality of basic kernel functions includes polynomial kernel functions of five different orders d.
5. The method for detecting quality parameters of agricultural products based on fluorescence hyperspectral imaging according to claim 1, wherein the agricultural products include kiwi fruits.
6. The method for detecting quality parameters of agricultural products based on fluorescence hyperspectral imaging according to claim 2 wherein the average spectral data of the region of interest is specifically that ENVI software is used to take a selected region in the fluorescence hyperspectral image of the agricultural product as the region of interest and calculate the average spectral data of the region.
7. A method for detecting quality parameters of agricultural products based on fluorescence hyperspectral imaging as claimed in claim 3 wherein in step S2, the Detrending algorithm Detrending is used to remove noise from fluorescence hyperspectral images.
8. The method for detecting quality parameters of agricultural products based on fluorescence hyperspectral imaging as claimed in claim 6 wherein the selected area is 3/4.
9. A computer device, characterized in that it comprises a processor and a memory, in which a computer program is stored which, when loaded by the processor, performs the method according to any of claims 1-8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210077713.2A CN114720436B (en) | 2022-01-24 | 2022-01-24 | Agricultural product quality parameter detection method and equipment based on fluorescence hyperspectral imaging |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210077713.2A CN114720436B (en) | 2022-01-24 | 2022-01-24 | Agricultural product quality parameter detection method and equipment based on fluorescence hyperspectral imaging |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114720436A CN114720436A (en) | 2022-07-08 |
| CN114720436B true CN114720436B (en) | 2023-05-12 |
Family
ID=82236448
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210077713.2A Active CN114720436B (en) | 2022-01-24 | 2022-01-24 | Agricultural product quality parameter detection method and equipment based on fluorescence hyperspectral imaging |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114720436B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104931470A (en) * | 2015-06-02 | 2015-09-23 | 江苏大学 | Fluorescence hyperspectral technology-based pesticide residue detection device and method |
| CN111523542A (en) * | 2020-04-21 | 2020-08-11 | 中国农业大学 | Method for constructing regression prediction model of total bacterial count and method for detecting freshness of fish by using regression prediction model |
| CN113917003A (en) * | 2021-08-30 | 2022-01-11 | 浙江工业大学 | Method for detecting content of imperatorin in angelica dahurica decoction pieces |
Family Cites Families (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105158186B (en) * | 2015-09-17 | 2017-09-22 | 南京农业大学 | A kind of method detected based on high spectrum image to ternip evil mind |
| CN108344701A (en) * | 2018-02-02 | 2018-07-31 | 东北电力大学 | Paraffin grade qualitative classification based on hyperspectral technique and quantitative homing method |
| CN109187378B (en) * | 2018-10-17 | 2021-04-16 | 四川农业大学 | Hyperspectral image-based nondestructive detection method for content of soluble solids in kiwi fruit |
| CN109613022A (en) * | 2019-01-25 | 2019-04-12 | 华南农业大学 | A kind of method, apparatus and system of low latitude high-spectrum remote-sensing detection Citrus Huanglongbing pathogen |
| WO2020158107A1 (en) * | 2019-01-28 | 2020-08-06 | 日本たばこ産業株式会社 | Sample quality determining method using fluorescent image, program, and device |
| CN110553999A (en) * | 2019-07-04 | 2019-12-10 | 浙江大学山东工业技术研究院 | Citrus huanglongbing detection method based on hyperspectral and chlorophyll fluorescence imaging fusion |
| CN112309498B (en) * | 2020-12-31 | 2021-04-09 | 武汉纽福斯生物科技有限公司 | Gene detection method and device based on deep learning and fluorescence spectrum |
| CN113008805B (en) * | 2021-02-07 | 2024-03-26 | 浙江工业大学 | Radix angelicae decoction piece quality prediction method based on hyperspectral imaging depth analysis |
| CN113496486B (en) * | 2021-07-08 | 2023-08-22 | 四川农业大学 | Kiwi fruit shelf life rapid discrimination method based on hyperspectral imaging technology |
| CN113820300A (en) * | 2021-10-28 | 2021-12-21 | 南京农业大学 | Method for predicting growth of pseudomonas fluorescens in pork |
-
2022
- 2022-01-24 CN CN202210077713.2A patent/CN114720436B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104931470A (en) * | 2015-06-02 | 2015-09-23 | 江苏大学 | Fluorescence hyperspectral technology-based pesticide residue detection device and method |
| CN111523542A (en) * | 2020-04-21 | 2020-08-11 | 中国农业大学 | Method for constructing regression prediction model of total bacterial count and method for detecting freshness of fish by using regression prediction model |
| CN113917003A (en) * | 2021-08-30 | 2022-01-11 | 浙江工业大学 | Method for detecting content of imperatorin in angelica dahurica decoction pieces |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114720436A (en) | 2022-07-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zhang et al. | Wood defect detection method with PCA feature fusion and compressed sensing | |
| CN110852227A (en) | Hyperspectral image deep learning classification method, device, equipment and storage medium | |
| Le | Application of deep learning and near infrared spectroscopy in cereal analysis | |
| Bai et al. | Accurate prediction of soluble solid content of apples from multiple geographical regions by combining deep learning with spectral fingerprint features | |
| CN112052755B (en) | Semantic convolution hyperspectral image classification method based on multipath attention mechanism | |
| CN113030001B (en) | Fruit sugar degree detection method and system | |
| Sabih et al. | Utilizing explainable AI for quantization and pruning of deep neural networks | |
| Yuan et al. | Moldy peanuts identification based on hyperspectral images and Point-centered convolutional neural network combined with embedded feature selection | |
| KR101750520B1 (en) | Method for improving detection performance of hyperspectral target detection by using target signal exclusion | |
| CN107680081B (en) | Hyperspectral image unmixing method based on convolutional neural network | |
| CN115661069A (en) | Hyperspectral anomaly detection method and computer device | |
| Bhole et al. | A transfer learning-based approach to predict the shelf life of fruit | |
| CN102135496A (en) | Infrared spectrum quantitative analysis method and infrared spectrum quantitative analysis device based on multi-scale regression | |
| CN112926625B (en) | Deviation influence factor analysis method for satellite radiation data | |
| Liu et al. | An improved method for predicting soluble solids content in apples by heterogeneous transfer learning and near-infrared spectroscopy | |
| CN113903407A (en) | Component identification method, component identification device, electronic equipment and storage medium | |
| CN114720436B (en) | Agricultural product quality parameter detection method and equipment based on fluorescence hyperspectral imaging | |
| US20200279148A1 (en) | Material structure analysis method and material structure analyzer | |
| Gao et al. | Mass detection of walnut based on X‐ray imaging technology | |
| Verdú et al. | Laser scattering imaging combined with CNNs to model the textural variability in a vegetable food tissue | |
| CN116858789A (en) | Food safety detection system and method thereof | |
| CN113887656B (en) | Hyperspectral image classification method combining deep learning and sparse representation | |
| JP7203252B2 (en) | A Sensory Evaluation Method for Spectral Data of Mainstream Cigarette Smoke | |
| CN116026780A (en) | Method and system for online detection of coating moisture absorption rate based on series strategy wavelength selection | |
| Srivatsa et al. | Application of least square denoising to improve admm based hyperspectral image classification |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |